A user reliability responsibility allocation method and system based on isoelectric quantity-consequent load benchmark and equivalent electric quantity

Abstract

The application discloses a user reliability responsibility allocation method and system based on isoelectric quantity-consecutive load benchmark and equivalent electric quantity, obtains user load, thermal power unit parameters, forced outage rate, wind and light forecast output and error distribution; carries out security constrained unit commitment scheduling based on net load, determines unit state, output and maximum callable capacity, and calculates electric quantity shortage expectation value through probability convolution. The fixed generation side operation state and capacity probability distribution are obtained, under the total electric quantity conservation and classified load transfer constraint, the system benchmark load which is optimal in reliability and minimum in transferred electric quantity is obtained through two-stage optimization, and the benchmark curve is decomposed according to the user electric quantity proportion. According to the actual load state period unit EENS rate, the period risk weight is constructed, the deviation of the actual load of the user relative to the benchmark load is converted into the risk equivalent adjustment electric quantity, and the risk final equivalent electric quantity is synthesized with the actual electric quantity, thereby the system actual total electric quantity shortage expectation value is allocated, and the closure of the total user reliability responsibility and the system actual reliability risk is realized.

Description

Technical Field

[0001] This invention relates to the field of power technology, and in particular to a method and system for allocating user reliability responsibility based on an equal power load benchmark and equivalent power. Background Technology

[0002] With the continuous expansion of new energy power generation capacity and the increasing diversification of power load structure, the power system's supply-demand balance and capacity adequacy are affected by multiple factors, including the start-up and shutdown status of thermal power units, cross-period ramp-up capabilities, forced outage risks, and the uncertainty of wind and solar power output. Under conditions of high-proportion new energy integration, system reliability risk depends not only on the cumulative electricity consumption or maximum load in a single period during the user's study period, but also on the distribution pattern of user load in different periods, the direction of load transfer, and the degree of temporal matching between load and the random available capacity on the generation side. Users with the same electricity consumption during the same study period may have significantly different impacts on the system's power supply reliability due to differences in their electricity consumption periods. Existing reliability assessment methods mainly rely on the installed capacity or static available capacity of generating units, failing to fully consider the limitations of unit start-up and shutdown status, actual dispatch output, and cross-period ramp-up capabilities on the available capacity, thus making it difficult to accurately reflect the actual power supply capacity and its temporal changes under specific operating conditions.

[0003] In terms of reliability responsibility attribution and allocation, existing methods mostly use user electricity consumption, maximum load, system peak load period load contribution, or static capacity ratio as the basis for allocation. While allocation by electricity consumption is simple to calculate, it treats unit electricity consumption in different time periods as having the same reliability impact, failing to reflect the risk differences between time periods. Allocation by maximum load or peak load contribution mainly focuses on a few high-load periods, making it difficult to cover all time periods within the study period, and also failing to reflect the dynamic matching relationship between renewable energy output, thermal power operation status, and user load. Some methods further use marginal reliability indicators at load points to evaluate user responsibility, that is, using the local rate of change of EENS relative to load to decompose user responsibility. However, EENS with respect to load level is usually a convex piecewise linear function, and its marginal rate of change varies with load level and capacity status. Using only marginal factors at the actual load point or other single operating points mainly reflects the local risk characteristics near that point, making it difficult to fully describe the overall reliability impact corresponding to the time-series changes in user load. At the same time, existing methods generally lack a unified, objective load comparison benchmark with a clear reliability meaning. If we directly compare the actual load of each user, the influence of user electricity consumption scale and time-series behavior can easily become mixed.

[0004] Therefore, it is necessary to propose a method that can take into account the actual operating status and random available capacity characteristics of the generation side, establish an objective benchmark for comparison of equal power load, and realize responsibility tracing and allocation based on the reliability risk differences of user load at different time periods, so as to improve the temporal identification capability, interpretability and total consistency of reliability responsibility allocation. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a user reliability responsibility allocation method and system based on the equivalent electricity-load benchmark and equivalent electricity, to solve the problems of insufficient connection between reliability evaluation and actual unit operating status in existing methods, local linearization deviation in single-point marginal allocation, difficulty in closing user responsibility to the actual EENS of the system, and inability to effectively distinguish between electricity consumption scale responsibility and load timing behavior responsibility. This invention can simultaneously reflect the impact of user electricity consumption scale and load timing behavior on system reliability risk under unified generation-side operating conditions, providing a quantitative basis for user reliability responsibility identification, reliability assurance cost allocation, cost deduction, reliability improvement compensation, and demand response economic incentives.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a user reliability responsibility allocation method based on equal power consumption-load alignment benchmark and equivalent power consumption, comprising the following steps:

[0007] Step S1: Obtain the actual power consumption data of each user at a specified time granularity within the research period, and overlay it hourly to form the actual time-series total load curve of the system. Obtain the technical and economic parameters of each thermal power unit in the power system, and obtain the predicted output values ​​and prediction error probability distribution of each wind farm and photovoltaic power station at a specified time granularity.

[0008] Step S2: Based on the actual total load of the system and the predicted output of each wind farm and photovoltaic power station, calculate the net load of the system in each period of the study period. With the goal of minimizing the operating cost, start-up and shutdown cost and the cost of curtailment of renewable energy, execute the safety-constrained unit combination scheduling to obtain the start-up and shutdown status and output scheduling plan of each thermal power unit. Based on the start-up and shutdown status of the thermal power units, the output of the previous period, the installed capacity and the upward ramping ability, calculate the maximum callable capacity of each thermal power unit in each period.

[0009] Step S3: Based on the maximum callable capacity and forced outage rate of each thermal power unit in each time period, establish the probability distribution of available capacity in the two states of "available-outage" for the thermal power unit. Based on the probability distribution of predicted output and prediction error of each wind farm and photovoltaic power station, establish the probability distribution of output in multiple states for wind power and photovoltaic power in each time period. By recursively convolving the two states of thermal power units and the discrete convolution of the probability distribution of output in multiple states of wind power and photovoltaic power, obtain the probability distribution of total available capacity on the generation side in each time period. Based on this, calculate the expected energy not served (EENS) value in each time period and sum them up to obtain the total expected energy not served value of the system in the study period.

[0010] Step S4: Fix the probability distribution of the total available capacity of the generation side for each time period obtained in Step S3. According to the ranking of the actual total load of the system, classify each time period in the study period into peak, flat and valley periods. Set the upper limit of the total amount of electricity allowed to be transferred out and transferred in for peak, flat and valley periods respectively, and set the upper limit of the total amount of electricity transferred in the system in the study period to characterize the overall load transfer capability of the system. Under the condition that the total electricity consumption of the system is the same in the study period and the load transfer constraints are met, a two-stage optimization method is used to solve the system equal electricity reliability benchmark load curve. The first stage is based on the full-constraint unit combination scheduling results in Step S2. With the goal of minimizing the expected value of the total electricity shortage in the study period, the system benchmark load vector is used as the optimization variable to obtain the minimum expected value of the electricity shortage in the feasible region of load transfer and the set of the optimal load curves for reliability. The second stage, under the condition that the minimum expected value of the electricity shortage in the first stage remains unchanged, with the goal of minimizing the total amount of electricity transferred in the system, selects the system equal electricity reliability benchmark load curve with the smallest time sequence change relative to the actual total load curve of the system from the set of the optimal load curves for reliability.

[0011] Step S5: Calculate the electricity consumption of each user during the study period and its proportion to the total electricity consumption of the system during the study period, and decompose the system's equal power reliability benchmark load curve hourly according to the proportion to obtain the equal power-load benchmark curve of each user.

[0012] Step S6: Based on the expected power shortage value of each time period corresponding to the actual total load curve of the system obtained in Step S3 and the expected power shortage value of the total system power during the study period, determine the actual load state time period unit EENS rate and the average actual power EENS rate of the study period. Divide the actual load state time period unit EENS rate of each time period by the average actual power EENS rate of the study period to obtain the time period risk weight of each time period.

[0013] Step S7: Calculate the load deviation between the actual power consumption of each user in each time period and its equivalent power consumption minus the load reference power. Based on the positive or negative nature of the load deviation, determine each time period as higher than the reference, lower than the reference, or equal to the reference. Based on the time period risk weight, calculate the risk equivalent increase in power consumption during the time period higher than the reference and the risk equivalent reduction in power consumption during the time period lower than the reference. Accumulate these amounts over the study period and determine the risk equivalent adjustment power consumption for each user during the study period based on the difference between the risk equivalent increase in power consumption and the risk equivalent reduction in power consumption.

[0014] Step S8: Add the actual electricity consumption of each user during the research period to the risk equivalent adjustment electricity consumption during the research period to obtain the risk final equivalent electricity consumption of each user. Calculate the proportion of the risk final equivalent electricity consumption of each user to the total risk final equivalent electricity consumption of all users, and allocate the insufficient expected value of the total system electricity consumption during the research period obtained in step S3 according to this proportion to obtain the final reliability responsibility of each user.

[0015] In a preferred embodiment, step S2 specifically includes:

[0016] If we assume that the power consumption of each user is constant at a given time granularity, then the total load of the power system in time period t is considered a fixed value. :

[0017]

[0018] Where T represents the total number of time periods in the research period; N represents the total number of users in the power system; Let d be the power consumption of user d during time period t.

[0019] During time period t, the predicted output values ​​for each wind farm and photovoltaic power station are: , The total load of the power system in time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t before the dispatch of thermal power units. This constitutes the net load curve at a specified time granularity within the power system research cycle:

[0020]

[0021] Let the initial start-up state and initial output of thermal power unit j in the period before the start of the study cycle be known quantities. and For the M thermal power units in the power system, a safety-constrained unit combination scheduling is implemented within the research period. The objective function is to minimize the sum of thermal power generation costs, unit start-up and shutdown costs, and renewable energy curtailment penalty costs within the system research period.

[0022]

[0023] in, ; The unit power generation cost of each thermal power unit obtained in step S1; The duration of a time period at a specified time granularity; The output of thermal power unit j during time period t; The starting cost of thermal power unit j; The downtime cost of thermal power unit j; Let j be the starting state variable of thermal power unit j in time period t. When, it indicates that thermal power unit j starts during time period t. When, it indicates that thermal power unit j was not started during time period t; Let j be the shutdown state variable of thermal power unit j during time period t. When, it indicates that thermal power unit j is shut down during time period t. When, it indicates that thermal power unit j did not shut down during time period t; Let be the wind curtailment power of wind farm s during time period t; Let g be the amount of curtailed solar power from photovoltaic power station g during time period t; This represents the penalty coefficient for abandoning renewable energy.

[0024] Define the start-stop state variable of thermal power unit j in time period t as follows: ,when The time indicates that thermal power unit j is in the operating state during time period t. Let t represent the time period in which thermal power unit j is in a shutdown state. The operating status, startup status, shutdown status, output scheduling plan, and renewable energy curtailment power of each thermal power unit in each time period within the study period are obtained by solving the safety-constrained unit combination model. Furthermore, considering the flexibility constraints of thermal power, the maximum available capacity of each thermal power unit in each time period is calculated based on the output of the thermal power unit in the previous time period, the operating status, startup status, installed capacity, and ramp-up capability of the current time period.

[0025]

[0026] in, Let j be the installed capacity of thermal power unit j; The output scheduling plan for thermal power unit j in time period t-1; This represents the operating status of thermal power unit j during time period t-1. The upward climbing ability of thermal power unit j within a specified time granularity; Used to characterize the startup status of thermal power units during time period t.

[0027] In a preferred embodiment: system operating constraints include:

[0028] System power balance constraints:

[0029]

[0030] Constraints on Curtailed Power from New Energy Sources:

[0031]

[0032]

[0033] Among them, the curtailed wind power shall not exceed the predicted output of the wind farm for the corresponding period, and the curtailed solar power shall not exceed the predicted output of the photovoltaic power station for the corresponding period.

[0034] Unit output upper and lower limit constraints:

[0035]

[0036] in, For the minimum technical output of thermal power unit j, To maximize the technical output of thermal power unit j;

[0037] Unit start-up and shutdown logic constraints:

[0038]

[0039]

[0040] in, and These represent the start-up and shutdown states of thermal power unit j in two adjacent time periods; when and Sometimes, , This indicates that the unit starts during time period t; and Sometimes, , This indicates that the unit is shut down during time period t; when the unit's start-up / shutdown status remains unchanged... and ;constraint This means that the same unit cannot be started and stopped simultaneously at the same time.

[0041] Unit ramp-up constraints:

[0042]

[0043] in, , These represent the upward and downward climbing capabilities of thermal power unit j within a specified time granularity, respectively, P j,t-1The output of thermal power unit j in time period t-1; the above constraints limit the output variation between adjacent time periods when the unit is continuously running; when the unit starts or stops, the output is controlled by the start-up state variable. and shutdown state variables Relaxing the output variation restrictions for the corresponding time periods allows the unit start-up and shutdown process to be correctly described by the unit combination model.

[0044] Spinning reserve constraints include upward spinning reserve and downward spinning reserve. The upward and downward spinning reserves available for thermal power unit j during time period t are defined as follows: , For thermal power units to be rotated upwards for standby, both the unit capacity margin constraint and the upward ramp accessibility constraint must be met simultaneously.

[0045]

[0046] The aforementioned upward rotation reserve constraint indicates that the upward reserve available by the unit cannot exceed the capacity margin between its current planned output and installed capacity, and also cannot exceed the maximum output increment it can achieve within a specified time granularity, starting from the previous operating point. Based on the definition of maximum available capacity in step S2, the upward rotation reserve can also be expressed as:

[0047]

[0048] in, Let be the maximum available capacity of thermal power unit j in time period t. Based on this, the upward rotation reserve can be calculated, which can simultaneously consider the upper limit of unit capacity and the operational accessibility constraint formed by the output of the previous time period and the upward climbing ability.

[0049] Thermal power units rotating downwards for standby must simultaneously meet both minimum technical output margin constraints and downward ramp accessibility constraints:

[0050]

[0051] The system satisfies the positive and negative rotating reserve requirements in time period t respectively:

[0052]

[0053] in, Given the maximum available capacity of thermal power unit j in time period t, the upward rotation reserve can be calculated, which can simultaneously consider the upper limit of unit capacity and the operational accessibility constraints formed by the output of the previous time period and the upward climbing ability. Minimum technical output for thermal power unit j; , These are the positive and negative spinning reserve requirements of the system in time period t, respectively. In this embodiment, both the upward and downward spinning reserve requirements of the system in time period t are set to 10% of the actual total load of the system in that time period.

[0054] Accordingly, the system's positive rotation standby constraint can be equivalently represented as:

[0055]

[0056] Minimum continuous uptime constraint for the unit:

[0057]

[0058] in, Let j be the minimum number of consecutive operating periods for thermal power unit j; when When, it indicates that thermal power unit j starts during time period t; when At that time, the constraint requires the unit to be continuously running from time period t. A time period; when At that time, this constraint does not impose additional restrictions on subsequent time periods. When the start period is close to the end of the research cycle, i.e. At that time, the above constraints ensure that the unit continues to operate for the remaining period of the study cycle.

[0059] Minimum continuous downtime constraint for the unit:

[0060]

[0061] in, Let j be the minimum number of consecutive downtime periods for thermal power unit j; when When, it indicates that thermal power unit j is shut down during time period t; when At that time, the constraint requires the unit to be shut down continuously for at least 10 consecutive periods starting from time period t. A time period; when At that time, this constraint does not impose additional restrictions on subsequent time periods. When the downtime period is close to the end of the study cycle, i.e. At that time, the above constraints ensure that the unit remains shut down for the remaining period of the study cycle.

[0062] In a preferred embodiment, step S3 specifically includes:

[0063] Step S2 yields the start-up and shutdown status, output scheduling plan, and maximum available capacity of the thermal power unit during each period of the study cycle. From step S1, the forced outage rate of thermal power unit j is... Then its availability rate is Define random variables The available capacity of thermal power unit j in time period t, and the maximum callable capacity of the unit in time period t. When the value is greater than zero, a probability distribution model of available capacity in two states, "available" and "out of service," can be constructed:

[0064]

[0065] Among them, when When, it indicates that thermal power unit j is in an available state; when When the maximum available capacity of thermal power unit j is zero during time period t, the available capacity of the unit degrades to a zero-capacity deterministic state, i.e. .

[0066] Under the condition that the availability states of each power generation unit are independent, a probability distribution of the total available capacity on the power generation side of the system is constructed by discrete convolution. Definition This represents the probability that the total available capacity of j thermal power units after aggregation equals X before time period t. The initial probability distribution when the system does not contain any thermal power units is:

[0067]

[0068] Based on the probability distribution properties of the sum of independent random variables, the probability distribution of the total available thermal power capacity in the current system is as follows: The probability distribution of the total available capacity of the j-th thermal power unit is obtained by discrete convolution with the probability distribution of the available capacity of the j-th thermal power unit:

[0069]

[0070] in, Let be the set of available capacity states of thermal power unit j during time period t. Since the available capacity of thermal power unit j is a random variable... Only and Since the probabilities are not zero in both states, the above convolution formula can be simplified to the following recurrence relation:

[0071]

[0072] in, When the j-th thermal power unit is in an available state, the j-th thermal power unit is compared with the previous one. The total available capacity X is jointly contributed by the thermal power units of the two units; This indicates that when thermal power unit j is in a forced shutdown state, the total available thermal power capacity of the system still increases from the previous value. The probability contribution of the total available capacity X of the thermal power units. If the current target value of the total available capacity X of the system is less than the maximum callable capacity of thermal power unit j. Then the former Total available capacity of Taiwan thermal power units The value must be negative to satisfy the target value of X for the total available capacity of the current system, which violates the physical constraint that the available capacity of thermal power units is non-negative. Therefore, the probability of such an impossible event is 0. Thus, if... Then take .

[0073] After recursive convolution of M thermal power units, the probability distribution of the total available thermal power capacity in time period t is obtained:

[0074]

[0075] in, This represents the probability that the total available thermal power capacity in time period t is equal to X.

[0076] Considering the impact of prediction errors on the output of wind farms and photovoltaic power plants, based on the predicted output values ​​and prediction error probability distributions of each wind farm and photovoltaic power plant obtained in step S1 for each time period, and under the condition that the renewable energy output is non-negative and does not exceed the corresponding installed capacity, a time-by-time multi-state available output probability distribution for each wind farm and photovoltaic power plant is constructed. For wind farm s, a random variable is defined. Let be the available power output of wind farm s during time period t, and its probability distribution be:

[0077]

[0078] in, Let be the discrete set of wind power output states of wind farm s during time period t, and satisfy . .

[0079] For a photovoltaic power station g, define a random variable. Let g be the available power output of the photovoltaic power station during time period t, and its probability distribution is as follows:

[0080]

[0081] in, Let g be the discrete set of photovoltaic output states of photovoltaic power station g during time period t, and satisfy the following conditions: .

[0082] Under the condition that the available power output of each wind farm and each photovoltaic power station is independent, the probability distribution of available power output of each wind farm and each photovoltaic power station within the same time period is discretized and convolved to obtain the probability distribution of total available renewable energy power output in time period t:

[0083]

[0084] Where, ∗ represents the convolution operation of two discrete probability distributions; This represents the probability that the total available output of renewable energy in time period t is equal to X.

[0085] Furthermore, by discretically convolving the probability distribution of total available thermal power capacity and the probability distribution of total available renewable energy output in time period t, the probability distribution of total available generation capacity in system time period t is obtained:

[0086]

[0087] Define the discrete state set of the total available capacity on the generation side for time period t as follows: Then we have:

[0088]

[0089] Probability distribution of total available generation capacity based on time period t Combined with the actual total load of the system Conduct a power supply reliability assessment. When the total available capacity X on the generation side is less than the total system load... At this time, the system has a power deficit, and the power deficit in this state is: When the total available capacity X on the generation side is not less than the total system load. At this time, the system has no power deficit; the power deficit is zero. Under a specified time granularity, the total power deficit of the entire system in time period t is less than the expected value EENS. t This is the weighted sum of the power deficit under all possible power outage conditions during the period and their probability of occurrence, multiplied by the specified duration of the period:

[0090]

[0091] in, The duration is specified for a given period of time;

[0092] The expected power shortage value EENS for each period of the power system during the study period. t Summing these values ​​yields the expected total energy deficit EENS over the research period. total :

[0093]

[0094] Among them, EENS t EENS represents the expected value of insufficient power during time period t. total The total power consumption is less than the expected value during the study period; T is the total number of time periods in the study period.

[0095] In a preferred embodiment, step S4 specifically includes:

[0096] The start-up, shutdown, output scheduling plan, and maximum available capacity of the thermal power units obtained in step S2 are fixed, as are the probability distributions of the total available capacity on the generation side for each time period obtained in step S3. During the system's equal-power reliability benchmark load optimization process, unit combination scheduling is not re-executed, nor is the probability distribution of the total available capacity on the generation side changed. This ensures that the actual system load curve and different candidate benchmark load curves are compared for reliability under the same generation side operating conditions and random available capacity. Let the random variable of the total available capacity on the generation side for time period t be... Its discrete state set is The probability that the total available capacity on the power generation side is equal to X during time period t is: And satisfy For any non-negative load level in time period t. Define the expected power shortage function for this period as follows:

[0097]

[0098] in, ; The duration of a time period at a specified time granularity.

[0099] Let the actual total load curve of the system during the study period be... Then the system load vector The corresponding expected value for total battery power deficiency is:

[0100]

[0101] The total power consumption during the system research cycle is:

[0102]

[0103] Based on the actual total system load Sort all time periods in ascending order, and then prioritize those with lower loads after sorting. The time period is divided into valleys. The higher load The time period is divided into peak periods. The remaining time periods are divided into flat periods. The actual total electricity consumption corresponding to peak, flat, and valley periods are respectively... , and Let the reference load curve for the system's isoelectric reliability be denoted as . Introduce nonnegative variables respectively and Let represent the load transfer-in power and load transfer-out power during time period t. Then, the baseline load to be optimized and the actual load satisfy the following condition: .

[0104] To characterize the overall load transfer capacity during peak, flat, and valley periods respectively, the proportions of the total electricity allowed to be transferred out and transferred in during peak periods are set as follows: and The proportions of total electricity allowed to be transferred out and transferred in during the flat section are respectively and The proportions of total electricity allowed to be transferred out and transferred in during off-peak hours are respectively and and satisfy .

[0105] The total electricity transferred out and transferred in during peak periods respectively meets the following requirements:

[0106]

[0107] The total electricity transferred out and transferred in during the flat section of the load respectively meets the following requirements:

[0108]

[0109] The total electricity transferred out and transferred in during off-peak hours respectively meets the following requirements:

[0110]

[0111] Furthermore, the total transferred electricity of the system during the study period is:

[0112]

[0113] The maximum transferable power during the system research cycle is set as the proportion of the total power consumption during the system research cycle. ,in Then the total transferred power of the system should meet the following requirements. In the formula, the coefficients This is used to avoid repeatedly calculating the same load transfer power.

[0114] To incorporate the expected function of insufficient power into the linear optimization model, a non-negative auxiliary variable is introduced. Indicates the baseline load to be optimized The power deficit under the total available capacity X of the generation side in time period t, satisfying:

[0115]

[0116] Therefore, the baseline load vector of the system to be optimized The corresponding expected value of total power consumption during the research period can be equivalently expressed as:

[0117]

[0118] Under the constraints described above, a two-stage optimization method is used to solve for the system's equal-power reliability baseline load curve. The first stage aims to minimize the expected total power shortage during the study period.

[0119]

[0120] In the first phase, there may be multiple candidate baseline load curves with the same minimum total power shortage expectation. To avoid unnecessary load rearrangement without further reliability improvements, the second phase retains all constraints from the first phase and adds constraints to keep the minimum total power shortage expectation unchanged from the first phase:

[0121]

[0122] Under these conditions, the second-stage objective is to minimize the total transferred power of the system.

[0123]

[0124] Let the optimal load transfer-in power and load transfer-out power in the second stage be respectively... and If the conditions are met simultaneously within the same time period and This can reduce both simultaneously. This process does not change the difference between the two, and therefore does not change the candidate baseline load and its total power deficiency expectation, but it will reduce the total transferred power of the system, which contradicts the optimality of the second-stage objective. Therefore, the optimal solution of the second stage must satisfy: Therefore, the optimal total transferred power in the second stage can be equivalently expressed as:

[0125]

[0126] After solving the above two-stage optimization model, the system's equivalent electrical reliability baseline load curve is obtained:

[0127]

[0128] In a preferred embodiment, step S5 specifically includes:

[0129] Furthermore, based on the principle of equal power consumption and load following, the system's equal power consumption reliability benchmark load curve is decomposed into equal power consumption and load following benchmark load curves for each user. The actual power consumption of user d during the study period is defined as:

[0130]

[0131] The total power consumption during the system research cycle is:

[0132]

[0133] Define the proportion of user d's electricity consumption during the research cycle to the total electricity consumption of the system during the research cycle as follows:

[0134]

[0135] Then the equivalent power consumption minus the load baseline power for user d during time period t is:

[0136]

[0137] Therefore, the equal power consumption-load baseline curves of each user satisfy the aggregation closure relationship in each time period:

[0138]

[0139] And ensure that the power consumption of each user remains constant throughout the research cycle:

[0140]

[0141] Meanwhile, the equal-load-progression baseline curves for each user and the system equal-load reliability baseline load curve have the same time-series shape:

[0142]

[0143] Therefore, the equal-load baseline curves of different users are scaled only according to the proportion of their respective power consumption in the study period to the total power consumption in the system study period, and their time-series shape remains consistent.

[0144] In a preferred embodiment, step S6 specifically includes:

[0145] Step S3 yields the expected power shortage value for time period t under the actual total load curve of the system. and the total power consumption during the research period is less than the expected value Based on the actual system load power consumption during time period t and its corresponding expected power deficit, the actual load state time period unit EENS rate for time period t is defined as follows:

[0146]

[0147] The actual load state time period unit EENS rate is used to characterize the average power shortage risk corresponding to the unit power consumption of the actual total load of the system in time period t under the conditions of the power generation side operating status determined in step S2 and the probability distribution of the total available capacity of the power generation side determined in step S3.

[0148] Based on the expected value of total power consumption during the system research cycle Total electricity consumption The average actual unit electricity consumption rate (EENS) during the research period is defined as:

[0149]

[0150] exist Under these conditions, the time period risk weight for time period t is obtained by dividing the actual load state time period unit EENS rate by the average actual actual unit electricity EENS rate of the study period:

[0151]

[0152] When the total power consumption of the system during the study period is less than the expected value of zero, the final reliability responsibility of each user is zero.

[0153] In a preferred embodiment, step S7 specifically includes:

[0154] The actual power consumption of user d during time period t is: Its equivalent power-load baseline power in time period t is Then, the load deviation of user d in time period t relative to the equal-load-follow-load benchmark is:

[0155]

[0156] Based on the sign of the load deviation, the time periods of user d within the study period are divided into sets higher than the baseline time periods. Sets below the baseline period The sum equals the base time set The three time periods are independent of each other but together cover all time periods within the research period. For any user and any time period, only one of the following states can be represented: above the benchmark, below the benchmark, or equal to the benchmark.

[0157] for If user d's actual power consumption in time period t is higher than its equivalent power consumption minus the load baseline power, then the risk weight for the unified time period obtained in step S6 is applied. Calculate the risk-equivalent increase in electricity consumption during this period:

[0158]

[0159] for If user d's actual power consumption during time period t is lower than its equivalent power consumption minus the load baseline power, calculate the risk-equivalent power reduction for that time period:

[0160]

[0161] for The user's actual power consumption is equal to its equivalent power consumption minus the load baseline power, and it does not result in a risk-equivalent increase in power consumption or a risk-equivalent reduction in power consumption during that period.

[0162] The risk-equivalent increase in electricity consumption for user d during all periods exceeding the baseline is summed to obtain the risk-equivalent increase in electricity consumption for the study period:

[0163]

[0164] The risk-equivalent electricity reduction for user d during all periods below the baseline is summed to obtain the risk-equivalent electricity reduction for the research period:

[0165]

[0166] The difference between the risk-equivalent increase in electricity consumption and the risk-equivalent reduction in electricity consumption during the research period is used to determine the risk-equivalent adjustment electricity consumption for user d during the research period.

[0167]

[0168] In a preferred embodiment, step S8 specifically includes:

[0169] User d's actual electricity consumption during the research period was: The equivalent adjusted electricity for the research cycle risk obtained in step S7 is Then the final equivalent electricity consumption for user d is:

[0170]

[0171] Based on the proportion of each user's final equivalent risk power to the total final equivalent risk power of all users, the total power consumption during the study period corresponding to the actual total load curve of the system calculated in step S3 is less than the expected value. Therefore, the final reliability responsibility of each user is:

[0172] .

[0173] This invention also provides a user reliability responsibility allocation system based on equal power consumption-load alignment benchmark and equivalent power consumption, including a processor, a memory, and a bus. The memory stores machine-readable instructions executed by the processor. The system is characterized in that, when the system is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the user reliability responsibility allocation method based on equal power consumption-load alignment benchmark and equivalent power consumption is as described above.

[0174] Compared with the prior art, the present invention has the following beneficial effects:

[0175] 1. This invention improves the consistency between reliability responsibility assessment and actual operating conditions. By determining the start-up and shutdown status and output plan of thermal power units through safety-constrained unit combinations, and comprehensively considering the unit ramp-up accessibility, forced shutdown risk, and uncertainty of renewable energy output, an hourly probability distribution of the total available capacity on the power generation side is constructed. Compared with methods using static installed capacity or fixed capacity margin, this method can more accurately reflect the power supply reliability risks under actual operating conditions.

[0176] 2. This invention establishes an objective and unified user load comparison benchmark. Under fixed generation side conditions, an equivalent power reliability benchmark load is obtained through two-stage optimization, and then decomposed according to the power consumption ratio of the user's research period. This allows each user to adopt the same benchmark time sequence shape while maintaining their own power consumption, thereby effectively distinguishing the impact of power consumption scale from the impact of load time sequence.

[0177] 3. This invention improves the consistency between user load time-series risk assessment and the actual system reliability status, and achieves closed-loop management of total system responsibility. By constructing the final equivalent electricity amount for risk using a unified time-period risk weight, user responsibility reflects the average reliability risk corresponding to their actual electricity consumption time series. Simultaneously, the actual total EENS of the system is allocated according to the proportion of the final equivalent electricity amount for risk, ensuring that the sum of all user responsibilities strictly equals the actual total EENS of the system. This provides a quantitative basis for reliability cost allocation and user load optimization. Attached Figure Description

[0178] Figure 1 Flowchart of the user reliability responsibility allocation method based on equal power load benchmark and equivalent power load in an embodiment of the present invention.

[0179] Figure 2 The actual time-series load, total system load, and predicted wind and solar power output and net load curves for each user at a 15-minute time granularity in this embodiment of the invention.

[0180] Figure 3 The predicted output values ​​and probability distributions of various wind farms and photovoltaic power stations at a 15-minute time granularity in this embodiment of the invention.

[0181] Figure 4 The present invention embodiment includes the hourly expected value of insufficient power corresponding to the actual total load of the system and the system's equivalent power reliability benchmark load, the unit EENS rate of the actual load state period, and the time period risk weight.

[0182] Figure 5 The equal power consumption-load baseline curves for each user in this embodiment of the invention.

[0183] Figure 6 The actual electricity consumption, risk-equivalent adjustment electricity consumption, and risk-final equivalent electricity consumption of each user in this embodiment of the invention.

[0184] Figure 7 The composition of the final reliability responsibility of each user in the embodiments of the present invention. Detailed Implementation

[0185] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0186] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0187] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application; as used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise; furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0188] refer to Figure 1-7 A user reliability responsibility allocation method based on equivalent power and load basis and equivalent power includes the following steps:

[0189] Step S1: Obtain the actual power consumption data of each user at a specified time granularity within the research period, and overlay it hourly to form the actual time-series total load curve of the system. Obtain the technical and economic parameters of each thermal power unit in the power system, and obtain the predicted output values ​​and prediction error probability distribution of each wind farm and photovoltaic power station at a specified time granularity.

[0190] Step S2: Based on the actual total load of the system and the predicted output of each wind farm and photovoltaic power station, calculate the net load of the system in each period of the study period. With the goal of minimizing the operating cost, start-up and shutdown cost and the cost of curtailment of renewable energy, execute the safety-constrained unit combination scheduling to obtain the start-up and shutdown status and output scheduling plan of each thermal power unit. Based on the start-up and shutdown status of the thermal power units, the output of the previous period, the installed capacity and the upward ramping ability, calculate the maximum callable capacity of each thermal power unit in each period.

[0191] Step S3: Based on the maximum callable capacity and forced outage rate of each thermal power unit in each time period, establish the probability distribution of available capacity in the two states of "available-outage" for the thermal power unit. Based on the probability distribution of predicted output and prediction error of each wind farm and photovoltaic power station, establish the probability distribution of output in multiple states for wind power and photovoltaic power in each time period. By recursively convolving the two states of thermal power units and the discrete convolution of the probability distribution of output in multiple states of wind power and photovoltaic power, obtain the probability distribution of total available capacity on the generation side in each time period. Based on this, calculate the expected energy not served (EENS) value in each time period and sum them up to obtain the total expected energy not served value of the system in the study period.

[0192] Step S4: Fix the probability distribution of the total available capacity of the generation side for each time period obtained in Step S3. According to the ranking of the actual total load of the system, classify each time period in the study period into peak, flat and valley periods. Set the upper limit of the total amount of electricity allowed to be transferred out and transferred in for peak, flat and valley periods respectively, and set the upper limit of the total amount of electricity transferred in the system in the study period to characterize the overall load transfer capability of the system. Under the condition that the total electricity consumption of the system is the same in the study period and the load transfer constraints are met, a two-stage optimization method is used to solve the system equal electricity reliability benchmark load curve. The first stage is based on the full-constraint unit combination scheduling results in Step S2. With the goal of minimizing the expected value of the total electricity shortage in the study period, the system benchmark load vector is used as the optimization variable to obtain the minimum expected value of the electricity shortage in the feasible region of load transfer and the set of the optimal load curves for reliability. The second stage, under the condition that the minimum expected value of the electricity shortage in the first stage remains unchanged, with the goal of minimizing the total amount of electricity transferred in the system, selects the system equal electricity reliability benchmark load curve with the smallest time sequence change relative to the actual total load curve of the system from the set of the optimal load curves for reliability.

[0193] Step S5: Calculate the electricity consumption of each user during the study period and its proportion to the total electricity consumption of the system during the study period, and decompose the system's equal power reliability benchmark load curve hourly according to the proportion to obtain the equal power-load benchmark curve of each user.

[0194] Step S6: Based on the expected power shortage value of each time period corresponding to the actual total load curve of the system obtained in Step S3 and the expected power shortage value of the total system power during the study period, determine the actual load state time period unit EENS rate and the average actual power EENS rate of the study period. Divide the actual load state time period unit EENS rate of each time period by the average actual power EENS rate of the study period to obtain the time period risk weight of each time period.

[0195] Step S7: Calculate the load deviation between the actual power consumption of each user in each time period and its equivalent power consumption minus the load reference power. Based on the positive or negative nature of the load deviation, determine each time period as higher than the reference, lower than the reference, or equal to the reference. Based on the time period risk weight, calculate the risk equivalent increase in power consumption during the time period higher than the reference and the risk equivalent reduction in power consumption during the time period lower than the reference. Accumulate these amounts over the study period and determine the risk equivalent adjustment power consumption for each user during the study period based on the difference between the risk equivalent increase in power consumption and the risk equivalent reduction in power consumption.

[0196] Step S8: Add the actual electricity consumption of each user during the research period to the risk equivalent adjustment electricity consumption during the research period to obtain the risk final equivalent electricity consumption of each user. Calculate the proportion of the risk final equivalent electricity consumption of each user to the total risk final equivalent electricity consumption of all users, and allocate the insufficient expected value of the total system electricity consumption during the research period obtained in step S3 according to this proportion to obtain the final reliability responsibility of each user.

[0197] A preferred embodiment is illustrated in the following specific example:

[0198] A 24-hour period was selected as the research period, with a 15-minute time granularity, divided into 96 time periods. Actual power consumption data for six typical user types were obtained: users 1 to 6, representing nighttime base load, all-day base load, daytime load, daytime-evening peak, morning peak, and evening peak users, respectively. Actual time-series load, total system load, predicted wind and solar power output, and net system load for each user are shown below. Figure 2 As shown in Table 1, the power generation side comprises 5 thermal power units with a total installed capacity of 7.80 MW. The main parameters of each thermal power unit are shown in Table 1. The minimum continuous start-up time and minimum continuous shutdown time for all units are set to 1 hour, i.e., 4 scheduling periods.

[0199] Table 1 Parameters of each thermal power unit

[0200]

[0201] The power system comprises four wind farms and two photovoltaic power stations, each with a rated capacity of 1 MW. The predicted power output and prediction error probability distributions for each wind farm and photovoltaic power station at a 15-minute time granularity are obtained as follows: Figure 3As shown in the figure. It is assumed that all units are already running before the start of the research period, but at minimum technical output. Therefore, the net system load is calculated based on the actual total system load and the predicted wind and solar power output. Safety-constrained unit combination scheduling is performed using the intlinprog solver in the MATLAB R2024a environment. Both upward and downward rotational reserve requirements are taken as 10% of the system load, and the renewable energy curtailment penalty coefficient is taken as 1 yuan / MWh. The maximum available capacity of each thermal power unit in each time period is obtained. Based on this, a probability distribution of the "available-out" two-state configuration of the thermal power units is constructed. Combined with the output probability distribution of each wind farm and photovoltaic power station, a probability distribution of the total available capacity on the generation side under a fixed operating benchmark is further constructed. The expected daily total power shortage value under the actual total load curve is calculated to be 0.3017 MWh.

[0202] Based on the ranking of the actual total system load from low to high, the 96 time periods were divided into 24 valley periods, 48 ​​flat periods, and 24 peak periods. The maximum total transferred electricity during the study period was set at 3% of the total system electricity consumption. The maximum allowed total transferred electricity during peak periods was set at 3% and 0.5% of the actual peak electricity consumption, respectively. The maximum allowed total transferred electricity during flat periods was set at 3% of the actual flat electricity consumption, and the maximum allowed total transferred electricity during valley periods was set at 0.5% and 5% of the actual valley electricity consumption, respectively. A two-stage optimization method was adopted, using the linprog solver to solve the system's equivalent electrical reliability baseline load curve. The corresponding time period EENS for the actual total system load and the system's equivalent electrical reliability baseline load were calculated. Simultaneously, the unit EENS rate and the average value over the study period were calculated for the actual load state time period, resulting in the time period risk weights. Figure 4 As shown. Based on the proportion of each user's electricity consumption during the research period to the total electricity consumption of the system during the research period, the system's equivalent electricity reliability baseline load curve is decomposed hourly, resulting in the equivalent electricity-load baseline curve for each user, as shown below. Figure 5 As shown.

[0203] Based on the deviation between each user's actual power consumption and the equivalent power consumption minus the load baseline, the risk-equivalent increase in power consumption and the risk-equivalent reduction in power consumption are calculated separately, and the difference between the two is used to obtain the risk-equivalent adjustment power consumption. The actual power consumption of each user during the study period is added to the risk-equivalent adjustment power consumption to obtain the final risk-equivalent power consumption. The actual power consumption, risk-equivalent adjustment power consumption, and final risk-equivalent power consumption for each user are as follows: Figure 6 As shown. The final reliability liability for each user, calculated based on the final equivalent electricity consumption ratio for risk, is 0.0083 MWh, 0.0761 MWh, 0.0678 MWh, 0.1234 MWh, 0.0045 MWh, and 0.0217 MWh, respectively, and its composition is as follows: Figure 7 As shown.

[0204] Users 1 and 2 have negative risk-equivalent adjustment amounts, indicating that their actual electricity consumption is more distributed in low-reliability-risk periods compared to the equal-load baseline, thus receiving corresponding risk-equivalent reductions. Users 3 to 6 have positive risk-equivalent adjustment amounts, indicating that their actual electricity consumption is relatively concentrated in higher-reliability-risk periods. User 4 has both a large electricity consumption during the study period and obvious daytime and evening peak electricity consumption characteristics, with a risk-equivalent adjustment amount of 19.4579 MWh and a final risk-equivalent amount of 58.6973 MWh, both the highest values ​​among the six user categories. User 2's actual electricity consumption is 44.2605 MWh, higher than User 4, but because its positive deviation from the equal-load baseline in high-risk periods is smaller, its risk-equivalent adjustment amount is -8.0503 MWh. These results show that the proposed method can further identify the matching relationship between user load timing and actual reliability risk based on the user's electricity consumption scale.

[0205] User 1's actual electricity consumption during the research period was 19.4763 MWh, but its load was mainly distributed during the low-risk period at night. The risk-equivalent adjustment amount was -15.5231 MWh, the final risk-equivalent amount was 3.9532 MWh, and the final reliability responsibility was 0.0083 MWh, significantly lower than the responsibility obtained by simply allocating according to the actual electricity consumption ratio. This indicates that the proposed method can avoid treating electricity consumption at different times as having the same reliability impact. Users 5 and 6 had actual electricity consumption of 2.0684 MWh and 2.0632 MWh, respectively, with basically the same scale. However, their risk-equivalent adjustment amounts were 0.0830 MWh and 8.2486 MWh, respectively. User 6's evening peak electricity consumption highly overlapped with the system's actual high-risk period; therefore, its risk-equivalent adjustment amount and final reliability responsibility were significantly higher than User 5's. This result shows that constructing time-period risk weights using the actual load state time-period unit EENS rate can enhance the identification ability of actual evening peak risks and their corresponding user load deviations.

[0206] User 6's evening peak electricity consumption overlaps more with high-risk periods than User 5's morning peak electricity consumption. Therefore, User 6's risk-equivalent adjustment electricity consumption and final reliability responsibility are both higher than User 5's. This result indicates that the proposed method can effectively identify the differences in reliability responsibility among users with similar electricity consumption but different electricity consumption periods.

[0207] Therefore, the method does not require the sum of the final equivalent electricity consumption of the risk to be equal to the actual total electricity consumption of the system. It can achieve the closure of the total actual reliability responsibility of the system based on the share of the final equivalent electricity consumption of the risk, and comprehensively reflect the differences between the user's electricity consumption scale during the research period, load timing deviation and actual time period reliability risk.

[0208] The above description is only a preferred embodiment of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention should be included in the scope of the present invention.

Claims

1. A user reliability responsibility allocation method based on equivalent power-load benchmark and equivalent power, characterized in that: Includes the following steps: Step S1: Obtain the actual power consumption data of each user at a specified time granularity within the research period, and overlay them hourly to form the actual time-series total load curve of the power system. Obtain the technical and economic parameters of each thermal power unit in the power system, and obtain the predicted output values ​​and prediction error probability distribution of each wind farm and photovoltaic power station at a specified time granularity. Step S2: Based on the actual total load of the power system and the predicted output of each wind farm and photovoltaic power station, calculate the net load of the system in each period of the study period. With the goal of minimizing the operating cost, start-up and shutdown cost and the cost of curtailment of renewable energy, execute the safety-constrained unit combination scheduling to obtain the start-up and shutdown status and output scheduling plan of each thermal power unit. Based on the start-up and shutdown status of the thermal power units, the output of the previous period, the installed capacity and the upward ramping ability, calculate the maximum callable capacity of each thermal power unit in each period. Step S3: Based on the maximum callable capacity and forced outage rate of each thermal power unit in each time period, establish the probability distribution of available capacity in the two states of "available-outage" for the thermal power unit. Based on the probability distribution of predicted output and prediction error of each wind farm and photovoltaic power station, establish the probability distribution of output in multiple states for wind power and photovoltaic power in each time period. By recursively convolving the two states of thermal power units and the discrete convolution of the probability distribution of output in multiple states of wind power and photovoltaic power, obtain the probability distribution of total available capacity on the generation side in each time period. Based on this, calculate the expected value of insufficient electricity EENS in each time period, and accumulate them to obtain the expected value of insufficient total electricity in the system within the study period. Step S4: Fix the probability distribution of the total available capacity of the generation side for each time period obtained in Step S3. According to the ranking of the actual total load of the system, classify each time period in the study period into peak, flat and valley periods. Set the upper limit of the total amount of electricity allowed to be transferred out and transferred in for peak, flat and valley periods respectively, and set the upper limit of the total amount of electricity transferred in the system in the study period to characterize the overall load transfer capability of the system. Under the condition that the total electricity consumption of the system is the same in the study period and the load transfer constraints are met, a two-stage optimization method is used to solve the system equal electricity reliability benchmark load curve. The first stage is based on the safety constraint unit combination scheduling results of Step S2. With the goal of minimizing the expected value of the total electricity shortage in the study period, the system benchmark load vector is used as the optimization variable to obtain the minimum expected value of the electricity shortage in the load transfer feasible region and the set of the optimal load curves for reliability. The second stage, under the condition that the minimum expected value of the electricity shortage in the first stage remains unchanged, with the goal of minimizing the total amount of electricity transferred in the system, selects the system equal electricity reliability benchmark load curve with the smallest time sequence change relative to the actual total load curve of the system from the set of the optimal load curves for reliability. Step S5: Calculate the electricity consumption of each user during the study period and its proportion to the total electricity consumption of the system during the study period, and decompose the system's equal power reliability benchmark load curve hourly according to the proportion to obtain the equal power-load benchmark curve for each user. Step S6: Based on the expected power shortage value of each time period corresponding to the actual total load curve of the system obtained in Step S3 and the expected power shortage value of the total system power during the study period, determine the actual load state time period unit EENS rate and the average actual power EENS rate of the study period. Divide the actual load state time period unit EENS rate of each time period by the average actual power EENS rate of the study period to obtain the time period risk weight of each time period. Step S7: Calculate the load deviation between the actual power consumption of each user in each time period and its equivalent power consumption minus the load reference power. Based on the positive or negative nature of the load deviation, determine each time period as higher than the reference, lower than the reference, or equal to the reference. Based on the time period risk weight, calculate the risk equivalent increase in power consumption during the time period higher than the reference and the risk equivalent reduction in power consumption during the time period lower than the reference. Accumulate these amounts over the study period and determine the risk equivalent adjustment power consumption for each user during the study period based on the difference between the risk equivalent increase in power consumption and the risk equivalent reduction in power consumption. Step S8: Add the actual electricity consumption of each user during the research period to the risk equivalent adjustment electricity consumption during the research period to obtain the risk final equivalent electricity consumption of each user. Calculate the proportion of the risk final equivalent electricity consumption of each user to the total risk final equivalent electricity consumption of all users, and allocate the insufficient expected value of the total system electricity consumption during the research period obtained in step S3 according to this proportion to obtain the final reliability responsibility of each user.

2. The user reliability responsibility allocation method based on equivalent power load and equivalent power as described in claim 1, characterized in that: The specific content of step S2 is as follows: Total load of the power system in time period t : Where T represents the total number of time periods in the research period; N represents the total number of users in the power system; Let d be the power consumption of user d during time period t; During time period t, the predicted output values ​​for each wind farm and photovoltaic power station are: , The total load of the power system in time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t before the dispatch of thermal power units. : Let the initial start-up state and initial output of thermal power unit j in the period before the start of the study cycle be known quantities. and The objective function is to minimize the sum of thermal power generation costs, unit start-up costs, unit shutdown costs, and renewable energy curtailment penalty costs within the system study period. A safety-constrained unit combination scheduling is implemented for M thermal power units in the power system within the study period. Based on the output of the thermal power units in the previous period, their current start-up status, startup status, installed capacity, and ramp-up capability, the maximum available capacity of each thermal power unit in each period is calculated. : in, Let j be the start-stop state variable of thermal power unit j during time period t. The time indicates that thermal power unit j is in the operating state during time period t. The time indicates that thermal power unit j is in a shutdown state during time period t; Let j be the installed capacity of thermal power unit j; The output scheduling plan for thermal power unit j in time period t-1; This represents the operating status of thermal power unit j during time period t-1; The upward climbing ability of thermal power unit j within a specified time granularity; Used to characterize the startup status of thermal power units during time period t.

3. The user reliability responsibility allocation method based on equal power consumption-load alignment benchmark and equivalent power consumption according to claim 1, characterized in that: The specific content of step S3 is as follows: Step S2 yields the start-up and shutdown status, output scheduling plan, and maximum available capacity of the thermal power unit during each period of the study period; step S1 determines the forced outage rate of thermal power unit j. Then its availability rate is Define random variables The available capacity of thermal power unit j in time period t, and the maximum callable capacity of the unit in time period t. When the value is greater than zero, a probability distribution model of available capacity in two states, "available" and "out of service," can be constructed: Among them, when When, it indicates that thermal power unit j is in an available state; when When the maximum available capacity of thermal power unit j is zero during time period t, the available capacity of the unit degrades to a zero-capacity deterministic state, i.e. ; Under the condition that the availability states of each power generation unit are independent, the probability distribution of the total available capacity of the system's power generation side is constructed by discrete convolution; [Definition] This represents the probability that the total available capacity of j thermal power units after aggregation equals X before time period t; the initial probability distribution when the system does not contain any thermal power units. for: The probability distribution of the total available thermal power capacity in the current system is as follows: The probability distribution of the total available capacity of the j-th thermal power unit is obtained by discrete convolution with the probability distribution of the available capacity of the j-th thermal power unit: in, When the j-th thermal power unit is in an available state, the j-th thermal power unit is compared with the previous one. The total available capacity X is jointly contributed by the thermal power units of the two units; This indicates that when thermal power unit j is in a forced shutdown state, the total available thermal power capacity of the system still increases from the previous value. The probability contribution of the total available capacity X of the thermal power units; if the current target value of the total available capacity X of the system is less than the maximum callable capacity of thermal power unit j. Then the former Total available capacity of Taiwan thermal power units The value must be negative to satisfy the target value of X for the total available capacity of the current system, which violates the physical constraint that the available capacity of thermal power units is non-negative. Therefore, the probability of such an impossible event is 0. Thus, if... Then take ; After recursive convolution of M thermal power units, the probability distribution of the total available thermal power capacity in time period t is obtained: in, This represents the probability that the total available thermal power capacity in time period t is equal to X; Based on the predicted output values ​​and prediction error probability distributions of each wind farm and photovoltaic power station obtained in step S1, a time-by-time multi-state available output probability distribution for each wind farm and photovoltaic power station is constructed. By recursively convolving the two states of thermal power units and discretely convolving the multi-state output probability distributions of wind power and photovoltaic power, the probability distribution of the total available capacity on the power generation side in each time period is obtained. Based on the probability distribution of the total available capacity on the power generation side, the expected value of insufficient power generation in each time period is calculated, and the total expected value of insufficient power generation in the system within the study period is accumulated.

4. The user reliability responsibility allocation method based on equivalent power-load benchmark and equivalent power as described in claim 1, characterized in that: The specific content of step S4 is as follows: The start-up status, startup status, shutdown status, output scheduling plan and maximum callable capacity of the thermal power unit obtained in step S2 are fixed, and the probability distribution of the total available capacity of the power generation side in each time period obtained in step S3 is also fixed. During the system's equal-power reliability benchmark load optimization process, the unit combination scheduling is not re-executed, nor is the probability distribution of the total available capacity on the generation side changed. This ensures that the actual system load curve and different candidate benchmark load curves are compared for reliability under the same generation-side operating conditions and random available capacity. Let the random variable of the total available capacity on the generation side in time period t be... Its discrete state set is The probability that the total available capacity on the power generation side is equal to X during time period t is: And satisfy For any non-negative load level in time period t Define the expected power shortage function for this time period. for: in, ; The duration of a time period at a specified time granularity; Let the actual total load curve of the system during the study period be... Then the system load vector The corresponding total power consumption is less than the expected value for: Among them, EENS t EENS represents the expected value of insufficient power during time period t. total To study the expected value of total power consumption within the research period; To incorporate the expected function of insufficient power into the linear optimization model, a non-negative auxiliary variable is introduced. Indicates the baseline load to be optimized The power deficit under the total available capacity X of the generation side in time period t, satisfying: Therefore, the baseline load vector of the system to be optimized The corresponding total power consumption during the research period is less than the expected value, which is equivalent. Represented as: Based on the actual total system load, all time periods are categorized into peak, flat, and valley time periods. Upper limits are set for the total electricity allowed to be transferred in and out during peak, flat, and valley periods, respectively, along with an upper limit for the total electricity transferred within the study period, to constrain the overall load transfer capacity of the system. Based on this, a two-stage optimization method is used to solve for the system's equivalent electricity reliability benchmark load curve. In the first stage, under the conditions of satisfying the load transfer constraints and the total electricity consumption conservation constraints of the study period, the objective is to minimize the expected value of the total electricity deficit corresponding to the candidate benchmark load during the study period. : in, The minimum expected total electricity shortage during the study period that can be achieved under given power generation side operating conditions, total available capacity probability distribution, and system load transfer capacity constraints. and These represent the load transfer-in power and load transfer-out power during time period t, respectively. Since the first-stage optimization model may have multiple candidate baseline load curves with the same minimum total energy deficit expectation, the second stage, while retaining all constraints of the first-stage optimization model and keeping the minimum total energy deficit expectation unchanged, aims to maximize load timing preservation, i.e., minimize the total system transferred energy. : The optimal candidate baseline load obtained in the second stage is determined as the system's equivalent electrical reliability baseline load curve. : 。 5. The user reliability responsibility allocation method based on equivalent power load benchmark and equivalent power as described in claim 1, characterized in that: The specific content of step S5 is as follows: Based on the principle of equal power consumption and load alignment, the system's equal power consumption reliability benchmark load curve is decomposed into equal power consumption and load alignment benchmark load curves for each user; the actual power consumption of user d during the study period is defined. for: The total power consumption E during the system research cycle is: Define the proportion of user d's electricity consumption during the research cycle to the total electricity consumption of the system during the research cycle as θ. d : Then the equivalent power consumption minus the load baseline power for user d during time period t is: Therefore, the equal power consumption-load baseline curves of each user satisfy the aggregation closure relationship in each time period: And ensure that the power consumption of each user remains constant throughout the research cycle: Meanwhile, the equal-load-progression baseline curves for each user and the system equal-load reliability baseline load curve have the same time-series shape: Therefore, the equal-load baseline curves of different users are scaled only according to the proportion of their respective power consumption in the study period to the total power consumption in the system study period, and their time-series shape remains consistent.

6. The user reliability responsibility allocation method based on equivalent power load and equivalent power as described in claim 1, characterized in that: The specific content of step S6 is as follows: Step S3 yields the expected power shortage value for time period t under the actual total load curve of the system. and the total power consumption during the research period is less than the expected value Based on the actual system load power in time period t and its corresponding expected power deficit, the actual load state time period unit EENS rate for time period t is defined as follows: The actual load state time period unit EENS rate is used to characterize the average power shortage risk corresponding to the unit power consumption of the actual total load of the system in time period t under the conditions of the power generation side operation status determined in step S2 and the probability distribution of the total available capacity of the power generation side determined in step S3. Based on the expected value of total power consumption during the system research cycle Total electricity consumption The average actual unit electricity consumption rate (EENS) during the research period is defined as: exist Under these conditions, the time period risk weight for time period t is obtained by dividing the actual load state time period unit EENS rate by the average actual actual unit electricity EENS rate of the study period: When the total power consumption of the system during the study period is less than the expected value of zero, the final reliability responsibility of each user is zero.

7. The user reliability responsibility allocation method based on equivalent power-load benchmark and equivalent power as described in claim 1, characterized in that: The specific content of step S7 is as follows: User d's actual power consumption during time period t is: Its equivalent power-load baseline power in time period t is Then, the load deviation of user d in time period t relative to the equal-load-follow-load benchmark is: Based on the sign of the load deviation, the time periods of user d within the study period are divided into sets higher than the baseline time periods. Sets below the baseline period The sum equals the base time set ; The three time period sets are independent of each other and together cover all time periods within the research period. For any user and any time period, only one of the following states can be corresponded: above the benchmark, below the benchmark, or equal to the benchmark. for If user d's actual power consumption in time period t is higher than its equivalent power consumption minus the load baseline power, then the risk weight for the unified time period obtained in step S6 is applied. Calculate the risk-equivalent increase in electricity consumption during this period: for If user d's actual power consumption during time period t is lower than its equivalent power consumption minus the load baseline power, calculate the risk-equivalent power reduction for that time period: for The user's actual power consumption is equal to its equivalent power consumption minus the on-load baseline power consumption, and it does not result in a risk equivalent increase in power consumption or a risk equivalent reduction in power consumption during that period. The risk-equivalent increase in electricity consumption for user d during all periods exceeding the baseline is summed to obtain the risk-equivalent increase in electricity consumption for the study period: The risk-equivalent electricity reduction for user d during all periods below the baseline is summed to obtain the risk-equivalent electricity reduction for the research period: The difference between the risk-equivalent increase in electricity consumption and the risk-equivalent reduction in electricity consumption during the research period is used to determine the risk-equivalent adjustment electricity consumption for user d during the research period. 。 8. The user reliability responsibility allocation method based on equal power consumption-load alignment benchmark and equivalent power consumption according to claim 1, characterized in that: The specific content of step S8 is as follows: User d's actual electricity consumption during the research period was: The equivalent adjusted electricity for the research cycle risk obtained in step S7 is Then the final equivalent electricity consumption for user d's risk is: Based on the proportion of each user's final equivalent risk power to the total final equivalent risk power of all users, the total power consumption during the study period corresponding to the actual total load curve of the system calculated in step S3 is less than the expected value. Therefore, the final reliability responsibility of each user is obtained as follows: 。 9. A user reliability responsibility sharing system based on equal power-load benchmark and equivalent power, comprising a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executed by the processor; characterized in that, When the system is running, the processor and the memory communicate via a bus, and the machine-readable instructions are executed by the processor as described in any one of claims 1 to 8, which is a user reliability responsibility sharing method based on equal power-load benchmark and equivalent power.

Images